摘要
本文研究了用遗传算法进行有限状态机 (FSM)的低功耗状态编码问题 .基于FSM的概率模型 ,对编码空间进行并行搜索 ;通过在适应性度量中引入面积和状态翻转信息 ,实现了面积和功耗之间的折衷 .对一些FSM的实际测试表明此方法平均能达到 2 0 %的功耗优化 .
This paper discusses about low power state encoding of finite state machines(FSM′s) using genetic algorithm(GA).Based upon the probabilistic model of FSM′s,GA based parallel searching is implemented in the encoding space.Moreover,area and state transition information may be both included in the ‘fitness’ measurement to realize some trade off between area and power.The test results for some FSM′s show that this approach may lead to 20% power reduction.
出处
《电子学报》
EI
CAS
CSCD
北大核心
2000年第8期124-126,共3页
Acta Electronica Sinica
基金
国家自然科学重大项目基金!(No .59995550 )
关键词
遗传算法
有限状态机
低功耗状态编码
genetic algorithm
finite state machine
low-power
encoding
fitness
